Is there a way to specify a number of IRISDB processes to be kept alive for serving requests and background jobs ?
In IRIS, every time a request need to be processed, a specific IRIS process (IRISDB.EXE) need to be assigned to handle that request.
If there is no spare IRIS process at that time, process will need to be created (and later destroyed). On some Windows systems (especially with security/antimalware solutions being active) creating new processes can be slow, which can results in delays during peak times (due to inrush of requests).
Because of that, it's sometimes useful to have some a certain number of processes already created and idling, for handling requests, especially during peak times (eg: in the morning). Is there a way to do that in IRIS ? (eg: in Portal or via config file)
I know it's already possible in Apache using the MinSpareServers setting and a specific NSD configuration.
The NSD process will hold connections made to IRIS, preventing IRISDB processes to be terminated once request is created.
That way it's possible to have a "pool" of IRISDB.exe processed to be created and stay.
However there is a major caveat which is those processes can be used for something else than a specific Apache node :
- If multiple Apache nodes are used (eg: using a load balancer), each of them will use N processes. For example with 4 Apache nodes each being set to hold 100 processes, 400 processes will be created. If one Apache node need more than 100 processes it cant use some from the 400 processes pool at IRIS side.
- Those processes can't be used for something else than requests (eg: jobs, IRIS work queue, ... ). Can be an advantage, depending how you look at it.
What I am looking is a setting, set at IRIS that will create and hold a certain amount of processes, regardless what they will be used for (either request or jobs).